Many researchers have suggested that local post-hoc explanation algorithms can be used to gain insights into the behavior of complex machine learning models. However, theoretical guarantees about such algorithms only exist for simple decision functions, and it is unclear whether and under which assumptions similar results might exist for complex models. In this paper, we introduce a general, learning-theory-based framework for what it means for an explanation to provide information about a decision function. We call an explanation informative if it serves to reduce the complexity of the space of plausible decision functions. With this approach, we show that many popular explanation algorithms are not informative when applied to complex decision functions, providing a rigorous mathematical rejection of the idea that it should be possible to explain any model. We then derive conditions under which different explanation algorithms become informative. These are often stronger than what one might expect. For example, gradient explanations and counterfactual explanations are non-informative with respect to the space of differentiable functions, and SHAP and anchor explanations are not informative with respect to the space of decision trees. Based on these results, we discuss how explanation algorithms can be modified to become informative. While the proposed analysis of explanation algorithms is mathematical, we argue that it holds strong implications for the practical applicability of these algorithms, particularly for auditing, regulation, and high-risk applications of AI.

In the rapidly growing literature on explanation algorithms, it often remains unclear what precisely these algorithms are for and how they should be used. We argue that this is because explanation algorithms are often mathematically complex but don't admit a clear interpretation. Unfortunately, complex statistical methods that don't have a clear interpretation are bound to lead to errors in interpretation, a fact that has become increasingly apparent in the literature. In order to move forward, papers on explanation algorithms should make clear how precisely the output of the algorithms should be interpreted. They should also clarify what questions about the function can and cannot be answered given the explanations. Our argument is based on the distinction between statistics and their interpretation. It also relies on parallels between explainable machine learning and applied statistics.

Users in many domains use machine learning (ML) predictions to help them make decisions. Effective ML-based decision-making often requires explanations of ML models and their predictions. While there are many algorithms that explain models, generating explanations in a format that is comprehensible and useful to decision-makers is a nontrivial task that can require extensive development overhead. We developed Pyreal, a highly extensible system with a corresponding Python implementation for generating a variety of interpretable ML explanations. Pyreal converts data and explanations between the feature spaces expected by the model, relevant explanation algorithms, and human users, allowing users to generate interpretable explanations in a low-code manner. Our studies demonstrate that Pyreal generates more useful explanations than existing systems while remaining both easy-to-use and efficient.

In this work, we explore various topics that fall under the umbrella of Uncertainty in post-hoc Explainable AI (XAI) methods. We in particular focus on the class of additive feature attribution explanation methods. We first describe our specifications of uncertainty and compare various statistical and recent methods to quantify the same. Next, for a particular instance, we study the relationship between a feature's attribution and its uncertainty and observe little correlation. As a result, we propose a modification in the distribution from which perturbations are sampled in LIME-based algorithms such that the important features have minimal uncertainty without an increase in computational cost. Next, while studying how the uncertainty in explanations varies across the feature space of a classifier, we observe that a fraction of instances show near-zero uncertainty. We coin the term "stable instances" for such instances and diagnose factors that make an instance stable. Next, we study how an XAI algorithm's uncertainty varies with the size and complexity of the underlying model. We observe that the more complex the model, the more inherent uncertainty is exhibited by it. As a result, we propose a measure to quantify the relative complexity of a blackbox classifier. This could be incorporated, for example, in LIME-based algorithms' sampling densities, to help different explanation algorithms achieve tighter confidence levels. Together, the above measures would have a strong impact on making XAI models relatively trustworthy for the end-user as well as aiding scientific discovery.

Artificial Intelligence (AI) is used in more and more high-stakes domains of our life such as justice [Berk, 2012], healthcare [Callahan and Shah, 2017], and finance [Lessmann et al., 2015], increasing the need to explain these decisions and to make sure that they are aligned with how we want the decision to be made. However, the complexity of many AI systems makes them challenging to comprehend, posing a significant barrier to their implementation and oversight [Arrieta et al., 2020, Samek et al., 2019]. Legislative initiatives, including the EU General Data Protection Regulation (GDPR), have recognized the'right for explanation' for individuals affected by algorithmic-decision making, emphasizing the legal necessity of explainability [Goodman and Flaxman, 2017]. In response, the field of Explainable Artificial Intelligence (XAI) has emerged, aimed at developing methods for explaining the decision-making processes of AI models [Adadi and Berrada, 2018, Holzinger et al., 2022, Xu et al., 2019]. Nevertheless, the landscape of post-hoc explanations is diverse, and each method can yield a different explanation. Furthermore, even within a single explanation method, multiple explanations can be generated for the same instance or decision. This phenomenon, known as the disagreement problem, has been studied in literature [Brughmans et al.,

Rule-based explanations provide simple reasons explaining the behavior of machine learning classifiers at given points in the feature space. Several recent methods (Anchors, LORE, etc.) purport to generate rule-based explanations for arbitrary or black-box classifiers. But what makes these methods work in general? We introduce a topological framework for rule-based explanation methods and provide a characterization of explainability in terms of the definability of a classifier relative to an explanation scheme. We employ this framework to consider various explanation schemes and argue that the preferred scheme depends on how much the user knows about the domain and the probability measure over the feature space.

Understanding the decision process of neural networks is hard. One vital method for explanation is to attribute its decision to pivotal features. Although many algorithms are proposed, most of them solely improve the faithfulness to the model. However, the real environment contains many random noises, which may leads to great fluctuations in the explanations. More seriously, recent works show that explanation algorithms are vulnerable to adversarial attacks. All of these make the explanation hard to trust in real scenarios. To bridge this gap, we propose a model-agnostic method \emph{Median Test for Feature Attribution} (MeTFA) to quantify the uncertainty and increase the stability of explanation algorithms with theoretical guarantees. MeTFA has the following two functions: (1) examine whether one feature is significantly important or unimportant and generate a MeTFA-significant map to visualize the results; (2) compute the confidence interval of a feature attribution score and generate a MeTFA-smoothed map to increase the stability of the explanation. Experiments show that MeTFA improves the visual quality of explanations and significantly reduces the instability while maintaining the faithfulness. To quantitatively evaluate the faithfulness of an explanation under different noise settings, we further propose several robust faithfulness metrics. Experiment results show that the MeTFA-smoothed explanation can significantly increase the robust faithfulness. In addition, we use two scenarios to show MeTFA's potential in the applications. First, when applied to the SOTA explanation method to locate context bias for semantic segmentation models, MeTFA-significant explanations use far smaller regions to maintain 99\%+ faithfulness. Second, when tested with different explanation-oriented attacks, MeTFA can help defend vanilla, as well as adaptive, adversarial attacks against explanations.

Existing and planned legislation stipulates various obligations to provide information about machine learning algorithms and their functioning, often interpreted as obligations to "explain". Many researchers suggest using post-hoc explanation algorithms for this purpose. In this paper, we combine legal, philosophical and technical arguments to show that post-hoc explanation algorithms are unsuitable to achieve the law's objectives. Indeed, most situations where explanations are requested are adversarial, meaning that the explanation provider and receiver have opposing interests and incentives, so that the provider might manipulate the explanation for her own ends. We show that this fundamental conflict cannot be resolved because of the high degree of ambiguity of post-hoc explanations in realistic application scenarios. As a consequence, post-hoc explanation algorithms are unsuitable to achieve the transparency objectives inherent to the legal norms. Instead, there is a need to more explicitly discuss the objectives underlying "explainability" obligations as these can often be better achieved through other mechanisms. There is an urgent need for a more open and honest discussion regarding the potential and limitations of post-hoc explanations in adversarial contexts, in particular in light of the current negotiations about the European Union's draft Artificial Intelligence Act.

The impressive performance of neural networks on natural language processing tasks attributes to their ability to model complicated word and phrase interactions. Existing flat, word level explanations of predictions hardly unveil how neural networks handle compositional semantics to reach predictions. To tackle the challenge, we study hierarchical explanation of neural network predictions. We identify non-additivity and independent importance attributions within hierarchies as two desirable properties for highlighting word and phrase interactions. We show prior efforts on hierarchical explanations, e.g. contextual decomposition, however, do not satisfy the desired properties mathematically. In this paper, we propose a formal way to quantify the importance of each word or phrase for hierarchical explanations. Following the formulation, we propose Sampling and Contextual Decomposition (SCD) algorithm and Sampling and Occlusion (SOC) algorithm. Human and metrics evaluation on both LSTM models and BERT Transformer models on multiple datasets show that our algorithms outperform prior hierarchical explanation algorithms. Our algorithms apply to hierarchical visualization of compositional semantics, extraction of classification rules and improving human trust of models.

Being able to reason about the predictions of a machine learning system is becoming increasingly important as sophisticated, non-linear predictive models are being adopted across the enterprise and beyond. In this talk we will discuss some requirements and challenges of model explanation algorithms and demo some practical examples using the open-source library Alibi we've developed at Seldon. - What makes an explanation interpretable? - The trade-off between interpretability and fidelity of an explanation algorithm - Practical examples of using some interpretable techniques (e.g.